What Eye Tracking Data Tells Us About Dyslexia: The Technology Behind the Technology
The D Labs team
Over the past four decades, scientists have become more and more aware of the fact that eye tracking methodologies are a valuable tool when it comes to investigating different aspects of human cognitive processing. While up until the last decade no one thought of using this tool to diagnose dyslexia, the high number of presently approximately one billion dyslexics around the world has made it become inevitable.
Dyslexic readers exhibit significant differences compared to normal readers with regard to reading eye-movement measures. It is quite common for children living with with dyslexia to complain about the words or lines moving in front of their eyes when reading, and for them to regularly skip over the smaller words such as is, if, at, a, I etc.
While reading a text, we make very abrupt jumps, called saccades, which take the gaze of our eyes to the next part of the text. During these fast abrupt movements our brains plug in the holes, as relatively no visual material is registered.
Amid these actions, the majority of word processing arises, the so-called fixations. Throughout a fixation we distinguish letter identities, classify the relative place of each letter and activate a word in our cerebral vocabulary that best matches the visual input. This entire process takes place within 200 milliseconds.
In reality, we actually skip over and not fixate at all on some words, as this makes our reading more efficient. On the other hand, when we are truly having trouble comprehending words, we often re-fixate on parts of text in order to re-read them. Tracking and evaluating eye movements – and in the process saccades, holes and fixations – provides us with a great opportunity to understand dyslexia.
Before going further, it is important to note that dyslexia is not a visual impairment or eye movement problem. It stems from difficulties in phonological processing and word decoding. These difficulties, however, are reflected in disruptive eye movement patterns during reading, which makes eye tracking a useful method for identifying individuals at risk.
Designing an eye tracking device specific for this need is the most complex part. There are two ways of designing an eye tracker:
1. A head mounted eye tracker
This is a system of measuring the angular gaze position of the human eye, by directing an infrared light beam along an optical axis into the eye, and displaying the reflected optical light distribution on a position sensing detector to determine eye position and create a heat map based on this.
It is a complex system head mounted on a child or adult during the reading time. Here is a sample design for the same on the left.
2. Several eye trackers mounted on a screen
A head mounted tracker has its own benefits. It is portable and thus able to track data at all times. However, it is hard to design well for kids with dyslexia. That is why a few of have put two to four eye trackers at each corner of a computing device. Screen size and positioning of the device is important here. This method allows high fidelity as well as ease of alignment and calibration.
A sample design for the same can be found on the top.
1. Heat maps tell us about the words a reader has concentrated on. For example, the following map shows that the reader has not concentrated enough on regular worlds such as “the” or “and”. This means that he or she has formed a blank image for those words. Moreover, after these words are encountered, the reader’s concentration increases by one time. That is why in this example, the dyslexic reader is concentrating twice as much on words such as “stone” and “fence” and three times as much on “through” and “pasture”.
2. The order that the reader has followed to read a sentence tells us about the image formation in their mind. When looking at the order of words read within the sentence it becomes clear that even though words such as “horse” and “stone have a higher concentration level, they were not re-read; telling us that the reader has already formed a mental image for the same.
Apart from letters, the experiment found, for instance, that developmental dyscalculia is associated with a subtilizing deficit, that is the inability to quickly enumerate small sets of up to three objects.
Eye tracking devices provide the latest piece in the puzzle for identifying and treating dyslexia.
The D Labs is an organization operating out of Bangalore, India with offices in San Francisco, USA and Berlin, Germany. It works in the areas of education technology and hardware. The D Labs have created a software-hardware combination to detect dyslexia at an early stage, for which they were awarded the Microsoft Imagine Cup Worldwide in 2012 as well as the UNICEF Global Health Award in 2013.